Intel announces First of its kind AI chips

Intel announced a set of AI (Artificial Intelligence) chips called Intel's Nervana Neural Network Processor lineup.

Training a network at a faster rate within a given power budget, accelerating deep learning deployment at scale and specialized leading-edge deep learning acceleration are some of the goals targetted with these AI chips.

Executive Opinion

Intel vice president and general manager, Artificial Intelligence Products Group, Naveen Rao, said, "To get to a future state of ‘AI everywhere,’ we’ll need to address the crush of data being generated and ensure enterprises are empowered to make efficient use of their data, processing it where it’s collected when it makes sense and making smarter use of their upstream resources. Data centers and the cloud need to have access to performant and scalable general-purpose computing and specialized acceleration for complex AI applications. In this future vision of AI everywhere, a holistic approach is needed—from hardware to software to applications."

The AI (Artificial Intelligence) chips

Intel's announcement noted that the dedicated accelerators for AI provide customers the right intelligence at the right time. Intel identified requirements for hardware architectures and complementary packaging, memory, storage and interconnect technologies to support different kinds of AI techniques

Code named Spring Crest, the first processor, Intel Nervana NNP-T (Neural Network Processor) built to prioritize and train a network as fast as possible and within the power budget.

Another processor code-named Spring Hill, Intel Nervana NNP-I built for inference. The processor uses Intel’s 10nm process technology with Ice Lake cores. The announcement claims that Spring Hill meets the performance per watt requirements for all major data center workloads. This chip is developed at Haifa, Israel development facility of Intel.

Reuters report also claimed the use of these AI chips at Facebook has already begun.